Professional Documents
Culture Documents
Quyet H. Cao剷 ک, Imran Khan ڨ, Reza Farahbakhsh ک, Giyyarpuram Madhusudan剷 ,
Gyu Myoung LeeƬ , Noel Crespiک
剷 Orange Labs, France, Email: ¢quyet.caohuu, giyyarpuram.madhusudan ¤ @orange.com
ڨSchneiderElectric Industries SAS, 38TEC, 38050, Grenoble Cedex 9, France, Email: imran@ieee.org
کInstitut Mines-Telecom, Telecom SudParis, CNRS UMR 5157, France,
Email: ¢reza.farahbakhsh, noel.crespi¤@it-sudparis.eu
Ƭ Department of Computer Science, Liverpool John Moores University, UK, Email: g.m.lee@ljmu.ac.uk
Abstract—The data generated by the devices and existing To deal with this issue of trust and control, we have proposed
infrastructure in the Internet of Things (IoT) should be a data usage control model to capture the diversity of
shared among applications. However, data sharing in the IoT obligations and constraints that data owners impose on the use
can only reach its full potential when multiple participants of data [4]. However, the architectural support to provide data
contribute their data, for example when people are able to use usage transparency and accountability is still lacking,
their smartphone sensors for this purpose. We believe that
motivating us to develop this type of architecture support
each step, from sensing the data to the actionable knowledge,
requires trust-enabled mechanisms to facilitate data exchange, for stakeholders in the context of shared platforms in smart
such as data perception trust, trustworthy data mining, and cities. The stakeholders themselves can thus participate in
reasoning with trust related policies. The absence of trust the sequence of steps in the mechanism that enhances the
could affect the acceptance of sharing data in smart cities. In transparency and accountability of data usage.
this study, we focus on data usage transparency and
We use the concept of ontologies and introduce the notion
accountability and propose a trust model for data sharing
in smart cities, including system architecture for trust-based of trust ontology, a formal representation of concepts related to
data sharing, data semantic and abstraction models, and a data usage control requirements, to annotate the data generated
mechanism to enhance transparency and accountability for by the devices or resources in smart cities. We have a semantic
data usage. We apply semantic technology and defeasible data model with which to present the number of entities, the
reasoning with trust data usage policies. We built a states of these entities. This leads to increased flexibility in
prototype based on an air pollution monitoring use case and terms of data integration, modeling, and processing compared
utilized it to evaluate the performance of our solution. to our previous data model based on NGSI [6]. This approach
is also aligned with the standardization reported in OneM2M
Keywords—Internet of Things, Smart Cities, Trust-based Data
Sharing, Data Usage Control, Defeasible Reasoning, and Air [7] as it provides the required abstractions.
Pollution Monitoring. Moreover, we provide trust enforcement for shared data
based on the consumers’ requests and policies of data owners,
allowing the IoT shared platform to keep track of data usage
I. I NTRODUCTION history. We then experiment further on a specific use case, using
a logic reasoner [8] to provide tests based on defeasible
Data sharing in the Internet of Things (IoT) [1] in general reasoning. Trust-based Data Usage (TDU) is the name of our
and in the context of smart cities [2] in particular will only reach solution.
its full potential if data can be collected by multiple sources.
One such example is that people are able to share their data The main contributions of this paper are four-fold: (i) A
related to different events by leveraging the sensing ca- multi-layer architecture for TDU - we describe a use case
pabilities of their smartphones. This crowd-sensing is a recent scenario, its background and main functional entities. We
trend [3] and may soon outperform traditional data collection also include a semantic and abstraction discussion for data
methods such as using pre-installed sensors. However, crowd- integration, modeling, and processing; (ii) A mechanism to
sensing may involve privacy issues for device owners. For ex- enhance the transparency and accountability of data usage - all
ample, some of the data collected by smartphones may contain the steps for stakeholders are provided; (iii) A TDU Ontology
sensitive information such as the location data of the owners. In (TDUO), created by extending some related concepts of the data
the context of smart cities, the data may come from a variety of usage conceptual model. We also define trust policies based on
sources, such as institutional actors, equipment manufacturers, defeasible rules and perform trust enforcement; and (iv) We
network operators, infrastructure providers, service providers, implement a prototype as a use case based on the TDU
and end users [4]. These data potentially undergo several architecture to evaluate its performance.
transformations, such as aggregation and composition, before The rest of the paper is organized as follows. Section
reaching their final destination. Another important aspect is II presents a motivating scenario to illustrate the need for
that the IoT data may also be shared for common usage through TDU. Section III presents our proposed system architecture in
linked data sets such as Linked Open Data [5]. Therefore detail and Section IV discusses the semantics and abstraction.
successful, and in some cases meaningful data sharing in smart Section V presents the transparency and accountability mecha-
cities depends on the establishment of trusted relationships nism. Section VI presents our prototype implementation along
among participants. We believe that participants will share their with the results. The related work is discussed in Section VII,
data when they have the ability to control the use of their data.
This paper has been accepted for publication in main track of IEEE International Conference on Communications 2016 (ICC 2016) to be held on 23-27 May 2016 at Kuala Lumpur Malaysia.
This is an author copy
C. Preliminary Performance Analysis Trust is the key for sharing IoT data among various
stakeholders. Using a simple scenario for smart cities, we
This part mainly presents the results based on tests for the propose a trust model to harmonize data sharing incorporating
trust enforcement time (TET) in the implemented architecture. policies defined by the data owner. In summary, we have
The goal is to show how much overhead is incurred due to the contributed a novel multi-layer architecture for TDU including
enforcement of data usage control based on the implementation a use case scenario, its background, main functional entities,
configuration in Section VI-B. and semantic and abstraction models. The mechanism for
To evaluate the proposed solution, we performed two transparency and accountability of data usage has provided
experiments and repeated each experiment 50 times. Their as a sequence diagram to the smart cities’ stakeholders. This
confidence interval is 95%. In the first experiment, the server also has proposed a TDUO trust ontology, defined trust policies
was restarted to create new server instance for each trust based on the trust ontology and defeasible rules, and performed
enforcement request. In the second experiment, we used the trust enforcement using defeasible reasoning. We finally have
same instance of server for subsequent trust enforcement implemented a prototype based on the TDU architecture to
requests. Based on that, we measured the TET in delay seconds. evaluate our solution.
Figure We learned several lessons from this study. The first is
6 shows the results of each experiment (with and without new that multiple stakeholders have to be involved in defining
server instance respectively). data sharing policies. In simple scenarios, these are normally
If server is restarted (new server instance is created for each defined by the data producer. However, in the IoT, data is
request) for each trust enforcement request we had average often merged from various sources and it becomes difficult to
delay of over 253ms. If same instance of server is used for determine who owns which data. The second lesson concerns
subsequent trust enforcement requests we have very less delay, the trust ontologies. We found that some effort needs to be
the average time was around 20ms. In most cases we did not expended in defining general and more open trust ontology,
have large delay, hence trust enforcement does not incur much a likely topic for future work. The third lesson is that the
additional delay. defining of trust and data usage policies should have input
from multiple actors. How data from IoT devices in a private
We also tried to use a cloud-based Google App Engine domain (e.g. smart homes) can be utilized and/or provided to
(GAE) but encountered exceptions such as access control interested entities needs to be explored. In some cases giving
exception while invoking the SPINdle logic reasoner. Solving incentives, like tax rebates, will be useful.
these exceptions involves deep understanding of GAE and For the next steps in this work, we will consider its
improve other performance analysis aspects, this is our planned implementation and validation in the real environment using
future work. end-to-end interactions. Another future work item is the
12 http://opdam2-opdam.rhcloud.com/
This paper has been accepted for publication in main track of IEEE International Conference on Communications 2016 (ICC 2016) to be held on 23-27 May 2016 at Kuala Lumpur Malaysia.
This is an author copy
development of an open source rule engine interpreter. This [17] J. Pato, S. Paradesi, I. Jacobi, F. Shih, and S. Wang, “Aintno: Demon-
rule engine can be used on the cloud-based GAE with no stration of Information Accountability on the Web,” in Privacy, Security,
Risk and Trust (PASSAT) and 2011 IEEE Third Inernational Conference
encountered exceptions and improved performance aspects. on Social Computing (SocialCom), 2011 IEEE Third International
Defeasible logic is a useful technique to check for inconsistent Conference on. IEEE, 2011, pp. 1072–1080.
rules, but we need to explore if there are other solutions available [18] M. Langheinrich, “Privacy by designprinciples of privacy-aware ubiq-
for this purpose. Lastly, we are going to provide a visualization uitous systems,” in Ubicomp 2001: Ubiquitous Computing. Springer,
tool to help users to customize their policies in an interactive 2001, pp. 273–291.
format that allows them to explore the consequences of certain [19] T. Kang and L. Kagal, “Enabling Privacy-Awareness in Social Net-
changes. works,” in AAAI Spring Symposium: Intelligent Information Privacy
Management, 2010.
[20] Z. Yan, P. Zhang, and A. V. Vasilakos, “A Survey on Trust Management
ACKNOW LEDGMENT for Internet of Things,” Journal of Network and Computer Applications,
The research leading to these results was partially funded vol. 42, pp. 120 – 134, 2014.
by the ITEA projects Fuse-IT, SEAS and CAP, and the ICT [21] P. Anantharam, C. Henson, K. Thirunarayan, A. P. Sheth et al., “Trust
model for semantic sensor and social networks: A preliminary report,”
R&D program of Information & Communications Technology in Aerospace and Electronics Conference (NAECON), Proceedings of the
Promotion (IITP) funded by the Korea government (MSIP) IEEE 2010 National. IEEE, 2010, pp. 1–5.
[R0190-15-2027, Development of TII (Trusted Information [22] G. Governatori, A. Rotolo, S. Villata, and F. Gandon, “One License
Infrastructure) S/W Framework for Realizing Trustworthy IoT to Compose Them All,” in The Semantic Web–ISWC 2013. Springer,
Eco-system]. 2013, pp. 151–166.
[23] A. Pretschner and T. Walter, “Negotiation of usage control policies-
R EFERENCES simply the best?” in Availability, Reliability and Security, 2008. ARES
08. Third International Conference on. IEEE, 2008, pp. 1135–1136.
[1] L. Atzori, A. Iera, and G. Morabito, “The Internet of Things: A Survey,” [24] A. Lazouski, F. Martinelli, and P. Mori, “Usage control in computer
Computer Networks, vol. 54, no. 15, pp. 2787–2805, 2010. security: A survey,” Computer Science Review, vol. 4, no. 2, pp. 81–
[2] A. Zanella, N. Bui, A. Castellani, L. Vangelista, and M. Zorzi, “Internet 99, 2010.
of things for smart cities,” Internet of Things Journal, IEEE, vol. 1,
no. 1, pp. 22–32, 2014.
[3] “Smart Urban Crowd-Sensing.” [Online]. Available: http://citris-
uc.org/infrastructure/project/smart-urban-crowd-sensing/
[4] Q. H. Cao, G. Madhusudan, R. Farahbakhsh, and N. Crespi, “Usage
Control for Data Handling in Smart Cities,” in IEEE GLOBECOM,
2015.
[5] T. Berners-Lee, “Linked Data,” in International Journal on Semantic
Web and Information Systems. W3C, 2006, vol. 4, no. 2.
[6] “NGSI 9/10 Information Model,” Tech. Rep. [Online]. Available:
http://goo.gl/mv6qFZ
[7] “OneM2M: Study of Abstraction and Semantics Enablements,” Tech.
Rep., 2015. [Online]. Available: http://goo.gl/2w98Y6
[8] H.-P. Lam and G. Governatori, “The making of SPINdle,” in Rule
Interchange and Applications. Springer, 2009, pp. 315–322.
[9] I. Khan, F. Belqasmi, R. Glitho, N. Crespi, M. Morrow, and P. Polakos,
“Wireless sensor network virtualization: early architecture and research
perspectives,” Network, IEEE, vol. 29, no. 3, pp. 104–112, May 2015.
[10] I. Khan, R. Jafrin, F. Errounda, R. Glitho, N. Crespi, M. Morrow, and
P. Polakos, “A data annotation architecture for semantic applications in
virtualized wireless sensor networks,” in Integrated Network Manage-
ment (IM), 2015 IFIP/IEEE International Symposium on, May 2015,
pp. 27–35.
[11] M. Compton, P. Barnaghi, L. Bermudez, R. Garcı́A-Castro, O. Corcho,
S. Cox, J. Graybeal, M. Hauswirth, C. Henson, A. Herzog et al., “The
SSN ontology of the W3C semantic sensor network incubator group,”
Web Semantics: Science, Services and Agents on the World Wide Web,
vol. 17, pp. 25–32, 2012.
[12] “RDF 1.1 Primer,” W3C, W3C Recommendation, 2014. [Online].
Available: http://www.w3.org/TR/rdf11-primer/
[13] S. Speiser, A. Wagner, O. Raabe, and A. Harth, “Web technologies and
privacy policies for the smart grid,” in Industrial Electronics Society,
IECON 2013-39th Annual Conference of the IEEE. IEEE, 2013, pp.
4809–4814.
[14] “ETSI M2M: Study on Semantic support for M2M data,” Tech. Rep.,
2012. [Online]. Available: http://goo.gl/LkYnsD
[15] O. Hartig, C. Bizer, and J. C. Freytag, “Executing SPARQL queries over
the web of linked data,” in The Semantic Web-ISWC, Springer Berlin
Heidelberg. W3C Working Group, 2009, pp. 293–309.
[16] D. Nute, “Handbook of Logic in Artificial Intelligence and Logic
Programming (Vol. 3),” D. M. Gabbay, C. J. Hogger, and J. A.
Robinson, Eds. New York, NY, USA: Oxford University Press, Inc.,
1994, ch. Defeasible Logic, pp. 353–395.